1. Field of the Invention
This invention generally relates to a system for performing the polymerase chain reaction (PCR) and more particularly to an apparatus and method for performing PCR on a DNA or RNA sample without removing it from its original cell structure.
2. Description of the Related Art
The polymerase chain reaction (PCR) is a process for making a very large number of faithful copies of a segment of double-stranded DNA (amplifying) by thermally cycling one or more molecules of this DNA (the template DNA) in the presence of thermally stable DNA polymerase enzyme, (typically Taq polymerase), the four DNA nucleotide bases and two or more single-stranded DNA primers. These primers are short segments of the order of 20 bases that are complementary to the 5' ends of the two complementary DNA single strands which make up the double-stranded template.
PCR is an immensely valuable technique which is very widely practiced, and has revolutionized the field of molecular biology. The technique is disclosed in detail in U.S. Pat. Nos. 4,683,195, 4,683,202, and 4,965,188. The reaction has until recently been carried out in solution in small reaction vessels, where the DNA to be amplified is in suspension. Apparatus for this process is disclosed in U.S. Pat. No. 5,038,852 and in U.S. patent application No. 07/871,274, filed Apr. 20, 1992.
Recently it has been found possible to apply PCR to amplify specific DNA segments inside cells, without first extracting the DNA from the cells. This technique,is called in situ PCR. The cells may be individual cells, or part of a tissue sample. Most often, in situ PCR is performed on cells or thin slices of tissue mounted on microscope slides. The cells or tissue usually have been fixed by treatment with formalin, or other reagents so that their morphology is preserved and recognizable after treatment.
If the selected DNA segment is amplified in such a way that the amplified product DNA can be selectively stained, then microscopic examination of the PCR treated cells can identify which cells in a tissue sample, if any, contain the specific DNA segment, and even where within the cell it is located. Visualization of the amplified DNA within the cells can be achieved by either of two methods. One method uses a complementary single stranded DNA probe to which a label molecule has been attached. This probe is hybridized to the specific DNA sequences in the amplified sample, if there are any, and the excess probe and label is washed away. Then the locations of the remaining label molecules are rendered visible by treatment with developer reagents. This technique for visualization is called "in situ hybridization", or "indirect detection" of the in situ amplified DNA.
The other method of visualization is to use, during the PCR process, a PCR reagent which includes modified DNA nucleotide bases to which a label molecule has been attached. Many modified bases carrying label molecules will be directly incorporated into the amplified product by the DNA polymerase. The location of the amplified DNA can then be visualized by treating the slide as before with developer reagents. This method is called "direct incorporation detection" of DNA amplified in situ.
The developer reagents in both methods typically include an enzyme such as alkaline phosphatase coupled to a molecule which binds strongly and specifically to the label molecule on the amplified DNA or on the hybridized probe, and a substrate which the enzyme converts into an insoluble, strongly absorbing dye. If the label molecule on the amplified DNA is biotin, then the binding molecule, coupled to the enzyme, is typically avidin. If the label molecule is digoxigenin, then the binding molecule is an anti-digoxigenin antibody. Both kinds of developing reagents have been used in both indirect detection by in situ hybridization and in direct incorporation detection. The labels on the label molecules may be colored, fluorescent, or radioactive.
Both methods of detection of in situ amplified DNA can be very sensitive, detecting from ten to a few hundred copies of the amplified DNA segment in each cell. Both require some post-PCR processing of the slide after the in situ PCR thermal cycling has been completed.
Methods for thermal cycling for in sitU PCR of whole cells differ depending on the cell source. Cells which do not form tissues, such as leukocytes and many cultured cells (such as HeLa cells), do not necessarily require special instrumentation for in situ PCR thermal cycling. Hasse et al, Proc. Natl. Acad. Sci. USA 87, 4971-4975 (1990) has reported that suspensions of such fixed cells can be thermal cycled in the same reaction tubes normally employed for solution phase PCR, and the cells subsequently spread on a slide for detection of the amplified product.
Often, however, such cells will be thermally cycled on a slide. They are spread out upon a slide by centrifugation, producing a so-called "cytospin" preparation. Spreading such cells facilitates fixation methods and other pre-PCR sample treatments. This does not require any additional effort, since as cytospin step would be otherwise be required later for visualization of the amplified product upon the slide. Cytospins are amplified identically to tissues mounted on slides.
When cells in tissue sections are to be studied, they must be amplified directly upon a surface, because the thin sections otherwise cannot display the tissue's morphology. The main problems associated with thermal cycling of these tissues are to maintain tissue morphology as well as cell morphology, inhibit evaporation of the PCR reaction mixture over the sample, and to obtain uniform, robustly repeatable results.
To perform in situ PCR on fixed cells or tissue samples on a glass microscope slide, one must first use a slide that has been treated so that the cells or tissue will stick to the slide and not be washed off or floated away by the aqueous reagents of the PCR process, or of the subsequent treatments for visualization of the amplified DNA. Typically a silanized slide, one treated so as to covalently bond 3-aminopropyl triethoxysilane molecules to its surface, works well. Alternatively, coatings of poly (lysine) or gelatin/chrome alum have been used.
Next, the area of the slide with the specimen to be amplified must be covered with an excess of the PCR reagent containing DNA polymerase enzyme, nucleotides, primers and other components at correct concentrations. Then the slide and reagents must be cycled typically 10 to 30 times between temperatures typically near 95.degree. C. and 68.degree. C., but sometimes as low as 37.degree. C., spending at least a fraction of a minute or more at each of two or three selected temperatures during each cycle.
There are several important requirements that must be met during thermal cycling for in situ PCR to be successful. One is that evaporation of water from the reagent must be almost completely prevented. No more than about 5% change from optimum reagent concentrations can be tolerated without lower amplification yields or poorer specificity resulting. Another requirement is that no material hostile to the PCR reaction may be in contact with the reagent during the process. Another is that bubbles of air or dissolved gas which are released by the reagent when it is heated should not disturb the access of the liquid reagent to the entire area to be processed.
Finally, published work on in situ PCR has shown that to preserve highly specific amplification it is often important to assemble the reaction so as to achieve a "hot start" or its chemical equivalent. In a physical hot Start, the complete reagent is neither assembled nor does it come into its first contact with the sample DNA until all the components required for the reaction are at a high temperature. The temperature must be high enough so that not even partial hybridization of the primers can occur at any locations other than the desired template location, in spite of the entire genome of the cell being available for non-specific partial hybridization of the primers. The temperature must also be high enough to prevent primer molecules from joining together to form an amplifiable product called "primer dimer". This safe starting temperature is typically in the range of 68.degree. C. to 75.degree. C., and typically is about 10.degree. C. hotter than the annealing temperature used in the PCR.
One way a "chemical hot start" in a PCR can be achieved is by including, in the reagent, a heat-labile component such as single strand binding protein (SSB) which prevents any extension by the polymerase enzyme until the reaction mixture has been heated in the first PCR cycle to a temperature high enough to prevent non-specific hybridization and also to destroy the heat-labile SSB component. Another way to implement a chemical hot start is to replace the dUTP with Uracil and to add to the reagent the heat-labile enzyme UNG (Uracil-N-Glycosylase) which destroys any PCR products made during a low temperature incubation prior to the first PCR cycle.
A still further way of implementing a chemical hot start is to combine the Taq polymerase enzyme with a Taq antibody before adding it to the reagent. Such a Taq monoclonal antibody has recently been announced by Dr. John Findlay of Kodak Clinical Products Division in a paper entitled "Development of PCR for In Vitro Diagnostics" presented at the 1992 san Diego Conference: Genetic Recognition, held Nov. 18-20, 1992. The Taq antibody binds to the Taq polymerase enzyme and inhibits its function at normal temperatures. However, upon heating the inhibited Taq polymerase to near 95.degree. C., the Taq antibody, a normal protein, is denatured, releasing the Taq polymerase enzyme and allowing it to function normally in the PCR process.
These types of chemical hot starts require inclusion of an often expensive component (e.g. UNG or Taq antibody) in the reagent and may place some undesirable constraints on the performance of the PCR, such as a relatively short time limit after a reagent is prepared before which it must be used, or a lower efficiency of amplification. Therefore, it is usually preferable to perform physical hot starts in in situ PCR if at all feasible.
Workers using in situ PCR today meet the requirements described above in a variety of ways that often they have developed in their own laboratories. The information they obtain using in situ PCR often cannot be obtained in any other way, and it is so valuable that they tolerate extraordinary inconvenience and investment of skill to get it.
A fairly common strategy has been to place a coverslip over a sample to be amplified and to seal it with nail polish or a similar adhesive as taught by Komminoth et al, Diagnostic Molecular Pathology 1, #2, 85-97 (1992). It is not unusual for such arrangements to leak, since the nail polish will not adhere strongly to tissue samples. Since all components containing the reagent are rigid, high pressures are produced at the denaturation temperature (typically 94.degree. C.) which can dislodge the nail polish. The hard cover slip can also damage the morphology of the fragile underlying cells if it touches them. This method does not allow for convenient hot starts. Another disadvantage is the requirement for a chloroform treatment of the assembly to dissolve the nail polish after cycling.
Such slides are generally laid upon the sample block of thermal cyclers that have not been explicitly designed for slides. To obtain good thermal contact, Komminoth et al used a spacer between the cycler's sample compartment cover and the slide to press the slide against the sample block. Further, existing thermal cyclers with holes in the thermal block accommodating reaction tubes do not have uniform thermal contact, and therefore, temperature uniformity for slides.
A related technique was demonstrated by Chiu et al, J. Histochemistry and Cytochemistry 40, #3, 333-341 (1992) who cultured cells to be studied on chambered slides, and applied in situ PCR to the cells upon the same slide. After culturing, the sample chamber walls were removed, but the gasket between the slide and chambers was left on the slide. In order to run an in-situ PCR, the cells were first covered with a PCR reaction mixture containing hot 2.5% agarose and the slides tightly wrapped with Saran wrap. This method provided a flexible cover for the cells, resting upon the gaskets which avoided injuring the cells. The Saran wrap provided evaporation control. The entire assembly was placed upon a Perkin-Elmer Cetus DNA Thermal Cycler with water added to the long slots between samples. The assembly was covered with plastic film and a plastic lid before cycling.
Nuovo et al, American Journal of Pathology 139, #6, 1239-1244 (1992) covered the samples with a coverslip made from flexible, temperature-stable polypropylene. The coverslip was typically anchored over the desired tissue section with a drop of nail polish at one corner. The slide is typically placed in an aluminum "boat" placed upon the sample block of a Perkin-Elmer Cetus DNA Thermal Cycler. This "boat" mainly serves the function to hold the mineral oil that was later added to the slide assembly to prevent evaporation of the reagents. After covering the sample with PCR reaction mixture without Taq DNA polymerase, the temperature of the sample was typically raised to about 65.degree., and the coverslip partially lifted to allow for the introduction of Taq DNA polymerase into the reaction mixture to initiate the PCR reaction.
The ability to lift up the coverslip allowed a hot start, which was demonstrated to produce major improvements in the specificity and yield of specific PCR product in the in situ PCR amplification. Following enzyme introduction, preheated mineral oil was added to the top and sides of the coverslip. Some of the oil was also drawn under the coverslip by capillary action. Its presence reduced the diameter of the droplet of PCR reagent under the coverslip, and also reduced control over the exact positioning of the droplet of PCR reagent. The method achieves excellent results, but requires one or more highly skilled operators, and is messy and unsuitable for large number of slides.
Another methodology employed to thermal cycle and reduce evaporation and condensation problems is to place slides within plastic bags and thermal cycle them in an air-oven type thermal cycler. For example, Staskus et al, Microbial Pathogenesis 1991, 11, 67-76 (1991) and Emberto et al, Proc. Natl. Acad. Sci. U.S.A. 90, 357-361 (1993) covered tissues with a coverslip and then mineral oil before placing the slides in a heat sealable plastic pouch cycled in an air-oven thermal cycler. The air oven approach avoids the possible problem of evaporation of water from the sample condensing on the cover slip because the entire assembly is roughly isothermal during thermal cycling.
The major disadvantage to the air oven approach is that poor heat transfer characteristics of the system results in very slow thermal cycling times. Also such systems often exhibit poor temperature uniformity from sample-to-sample in the oven.
In summary, none of the existing methods incorporate all the desirable aspects of good thermal uniformity, evaporation control without mineral oil, small reagent volume, maintenance of cell morphology, the ability to do hot starts, and convenience for assembling large numbers of slides for thermal cycling a position other than horizontal. Almost everyone today uses immersion of the prepared slide in mineral oil to prevent evaporation of water from the reagent during thermal cycling.
The oil overlay is a major inconvenience because usually it must be carefully and completely removed by additional steps after thermal cycling and before subsequent processing with detection reagents. Oil is also sometimes a carrier of contaminants which can kill the PCR.
We have found in experimental work that when containment of the aqueous-based reagent on the sample is achieved by mounting a plastic cover-slip over the specimen and attaching it to the slide by adhesives at the corners or edges, surface tension effects and elution of gas move the liquid reagent about during cycling. As a result, these methods are often not totally reliable. Generally, they require the slide to be horizontal, limiting the ways in which it can be held in a thermal cycler. Efforts to seal the reagent under a cover slip by completely surrounding its perimeter with a sealant adhesive often fail because a gas-tight seal is difficult or impossible to maintain with adhesives alone. In addition, the surface of the slide is usually coated with a thin layer of the specimen under at least part of the perimeter of the cover slip, so there is poor adhesion to the slide.
Accordingly there is a need for a new and improved complete sample containment system for performing in situ PCR. There is a need for a system which provides a convenient physical containment for the reagent physically against the sample which prevents reagent and sample evaporation during thermal cycling. There is also a need for a system which eliminates the need for the use of oil, which uses no adhesives, and does not require the slide to be horizontal during thermal cycling.